Gun Violence and Police Shooting in the U.S.

1 Introduction

The U.S. is an exceptional country when it comes to guns. It’s one of the few countries in which the right to bear arms is constitutionally protected. But the relationship between U.S. and guns is in a crucial way, which indicates it is far and away from the most homicidal incidents among all of the developed nations. Moreover, the causation of the recent well-known Black Lives Matter movement drives us to further relate U.S. gun incidents with police gun shootings. Thus, it led us curious about public safety and injustice in the United States since it is affecting all parts of our population, ourselves included.

The data in this dataset was downloaded from kaggle. In the following report, we will mainly use two datasets, US Police Shootings and US Gun Violence. US Gun Violence gathered by Eric Jing The CSV file contains data for all recorded gun violence incidents in the US between January 2013 to March 2018. The data consist of 240k incidents, with detailed information about each incident. US Police Shootings collected by Ahsen Nazir, a data analyst. The data is about police shootings in the United States across different states. The data also involves details about a race, whether the person was armed, threatening had a mental illness, etc. It contains 5924 observations and 15 features. One of the questions can be related to this dataset is whether or not specific racial groups are more subjected to police shootings.

2 Executive summary

The primary data we looked at (USPoliceShootings) has 4895 observations and 15 columns. The columns contain information about the id, name, date, manner of death, age, gender, race, city, state, signs of mental illness, threat level, flee, body camera, and arms category. We have looked at two liberal states (Washington and California) and two conservative states (Minnesota and Texas). The shooting incidents were 126, 701, 61, 426 in those states respectively. We think we also need population information for those states to make a more fair analysis about police violence in those states. The mean age of the victims was 36.5497496. The dataset contains data from 2015 to 2020, though only through June in 2020. The incidents were 965, 904, 906, 888, 858 and 374 respectively in those years. The dataset also contained information about 4 racial groups. The number of incidents for the Asian category was 93, 1298 for Black, 902 for Hispanic and 2476 for White racial group. 15 of the victims also showed signs of mental illness.

3 Table

  • We combine both US Police Shootings dataset and US Gun violence dataset by their mutual time range, which is from 01/02/2015 to 03/31/2018.
  • From dataset, we find out that gun death only holds little of the proportion among gun violence.
  • By combining both data, it is help us to understand that police shooting incidents are only occupy 2.0065361% among all gun violence in the U.S..
  • To further extend the understanding of the police shooting incidents, we include the race of the victims in the police shooting incidents to explore is that there exist racial issue and injustice.
## # A tibble: 188,185 x 6
## # Groups:   police_shooting_victims_race [7]
##    date   location   n_killed n_injured incident_characterist… police_shooting_…
##    <chr>  <chr>         <int>     <int> <chr>                  <chr>            
##  1 2015-… Norfolk, …        0         2 Shot - Wounded/Injured <NA>             
##  2 2015-… Norfolk, …        0         1 Shot - Wounded/Injured <NA>             
##  3 2015-… Virginia …        0         1 Shot - Wounded/Injured <NA>             
##  4 2015-… Surfside …        0         0 Shots Fired - No Inju… <NA>             
##  5 2015-… Jacksonvi…        0         1 Shot - Wounded/Injured <NA>             
##  6 2015-… Chicago, …        0         1 Shot - Wounded/Injured <NA>             
##  7 2015-… Savannah,…        1         4 Shot - Wounded/Injure… <NA>             
##  8 2015-… Chicago, …        0         2 Shot - Wounded/Injure… <NA>             
##  9 2015-… Saint Lou…        0         1 Shot - Wounded/Injure… <NA>             
## 10 2015-… Pittsburg…        1         0 Shot - Dead (murder, … <NA>             
## # … with 188,175 more rows

4 Data Exploration

4.2 US Police Shooting Victims by Race

  • This bar chart was intended to show trends in gun incidents in the U.S. from 2013 to 2018.
  • It shows the distribution of the race of each of the victims of a fatal police shooting.
  • The majority of victims have been white, almost double the amount of black victims which is the second highest.
  • Asian, Native, and other races have significantly smaller incidences.

4.3

  • This pie chart was intended to show the proportion of different characteristics of gun incidents in the U.S. from 2013 to 2018.
  • The top 7 high level characteristics that have the greatest rate of occurrence are “Shot-Wounded/Injured”, “Shot-Dead”, “Non-shooting Incident”, “Shots Fired - No injuries”, “Armed robbery”, “Non-shooting Incident”, “Institution/Group/Business”, and “TSA action”.
  • There are totally 53 types of high level characteristics. For a better visualization as a whole, rarer characteristics were grouped into a category named “other”, but it doesn’t mean the completely ignoring of those characteristics. In later data exploration, we will also zoom into this category.
  • More than half of the gun violence incidents were resulted in either injury or deaths.
  • When talking about gun violence, there are a also a lot of non-shooting incidents.